Abstract:  The data used in the model (such as temperature data) should be the actual environment data, however, most of the temperature data used in the previous crop model simulation studies are the observation temperature of local meteorological stations, i.e. the observation temperature in the shutter box, which is different from the actual air temperature in the rice field (at different heights). In order to study the difference between the actual temperature at different heights of the rice field and the station temperature and their effects on the simulation of the ORYZA2000 rice model, using the crop data and meteorological data from the 2016?2018 field trials of high thermal damage to rice at Shouxian Rice Test Station in Huainan City, Anhui Province, the ORYZA2000 rice model was calibrated locally, then the model was used to conduct comparative simulation research on the measured temperature in the rice field and the station temperature, at last, the differences between the actual temperature and the station temperature at different heights of rice field and their effects on the simulation results of ORYZA2000 rice model were analyzed. The research results show that: (1) the field air temperatures of 35, 75, and 125cm over the ground were obtained, and the field air temperature of each height were combined into the rice field temperature(called combined temperature) according to the measured plant height of the rice (which can represent the habitat temperature of the rice canopy in the field). The four temperatures and the station temperature are compared with each other: in terms of the average daily maximum temperature, the combined temperature of the rice field is the highest, which is nearly 2.5℃ higher than the station temperature; in terms of the average daily minimum temperature, the temperature gaps of each height in the field and the station temperature were relatively small.(2) After calibration of localized crop parameters, the ORYZA2000 rice model has higher accuracy in simulating rice growth period and dry biomass in Shouxian, and the simulation accuracy of crop growth period and total aboveground dry matter (WAGT) are very well. (3) In terms of rice yield simulation, the yield simulation effect using the rice field combined temperature is better than the station temperature, and the NRMSE of the rice field combined temperature and the station temperature simulated yield are 4.2% and 6.1%, respectively; research shows that using the actual field temperature in the ORYZA2000 rice model simulation is more reasonable than using the station temperature, especially when high thermal damage to rice is likely to occur.

Key words:  Anhui Province, Rice, Measured temperature in rice field, ORYZA2000 rice model, Yield simulation